Vantage Theory and Linguistics
Linguistic applications of Robert E. MacLaury's Vantage Theory

maintained by: Adam Głaz
UMCS, Lublin, Poland
[Rozmiar: 7350 bajtów]

[Rozmiar: 58591 bajtów]

Rob MacLaury's picture on his 53rd birthday in 1997
(received from Maru MacLaury)

Tribute to Robert E. MacLaury Obituary for R.E. MacLaury. Linguist List 15.1020, Mar 27, 2004. By Terri MacKeigan.

In Memory of Robert E. MacLaury, 1944-2004. Cross-Cultural Research, 39-1. 2005. (Special issue on Culture, Cognition and Color Categorization and Naming.) 3-4.

Dr Robert E. MacLaury 1944-2004: An Appreciation. By Terri MacKeigan and Chris Sinha. In Biggam, Carole P. and Christian J. Kay (eds.) 2006. Progress in Colour Studies. Volume I. Language and Culture. ix-x. Amsterdam/Philadelphia: John Benjamins.

Foreword: Dedication to Robert E. MacLaury. By Luisa Maffi. In Robert E. MacLaury, Galina V. Paramei and Don Dedrick (eds.) 2007. Anthropology of Color: Interdisciplinary Multilelvel Modeling.vii-ix. Amsterdam/Philadelphia: John Benjamins
Basic information on VT What is Vantage Theory?

Why Vantage Theory and linguistics?


Vantage Theory bibliography

Major publications on VT Color and Cognition in Mesoamerica, 1997, paperback 2011, Austin: University of Texas Press. [Rob's magnum opus]

Language Sciences 24 (5-6), 2002 (special issue; guest ed. Robert E. MacLaury)

Language Sciences 32 (2), 2010 (special issue; guest ed. Adam Głaz and Keith Allan)
Some unpublished papers (abstracts) MacLaury, Robert E. 1999. Vantage theory in outline. Unpublished Ms. (pdf)

MacLaury, Robert E. and Maria I. MacLaury. 2002. Vantages of a Song about Love. Unpublished letter. (pdf)

MacLaury, Robert E. 2003. Vantage Theory: Categorization as Space-time Analogy. Unpublished Ms. (pdf)
Recent Głaz, Adam. 2012. Extended Vantage Theory in Linguistic Application. The Case of the English Articles. Lublin: Wydawnictwo UMCS. [proposes an extension of VT as originally formulated and applies it to the intricacies of the use of the English articles]

Głaz, Adam, Marnie L. Moist and Elena Tribushinina. (eds.) forthcoming in 2013. Vantage Theory: A View on Cognition, Categorization and Language. [will contain so far unpublished work by Rob MacLaury as well as chapters dealing with VT vs. psychology, language, and categorization]

Many thanks to Maru MacLaury for all her help and support!

What is VT?

NOTE: Below is only a short and rather simplistic introduction to VT. To have a better understanding of the model, one must consult the VT publications (almost all Rob's works and many by other authors contain a preliminary section explaining the theory's basics). A very brief summary view is Keith Allan's entry on "Categorizing Percepts: Vantage Theory" in Encyclopedia of Languages and Linguistics (2nd edn), ed. by E. Keith Brown. Oxford: Elsevier. Perhaps Rob's most accessible introductory treatment of VT is his 2002 article "Introducing vantage theory" in Language Sciences, 24, 5-6.

At the most general level, VT holds that people categorize by performing a subconscious and instinctive analogy to the way they orient themselves in space-time. They plot their position relative to the spatial coordinates of up-down, left-right and front-back (combined into a unitary bodz of reference ) and the mobile coordinate of relative motion. However, a system of spatial coordinates can itself be moving, which affects judgment. MacLaury (e.g. 1995: 240; 1997: 143) quotes Einstein's classic example of a rock dropped from a moving train: its trajectory is different for a person on the train than for someone standing by the track.

By analogy, colour categorization involves combining into coherent wholes the fixed coordinates of hue, brightness or saturation with the mobile coordinates of reciprocally balanced degrees of attention, on the part of the viewer, to similarity or difference between colour stimuli. For example, the category RED is constructed by focusing on what the conceptualizer considers to be the "best example of red" and regarding other sensations as similar to or different from it. The stronger the attention to similarity, the more stimuli are included in the category so the range of the category expands. Attention to difference, in turn, results in marking the category's boundary: stimuli are deemed to lie outside the category so that the latter is curtailed at a margin.

Combining fixed and mobile coordinates into a coherent whole happens in at least two steps and is called vantage construction. Figure 1 models the abstract and imaginary category RED in a schematic way:

Rozmiar: 36408 bajtów

Figure 1. A schematic VT model of the category RED

First, the red hue (symbolized by R) is chosen as the primary fixed coordinate (the "starting point" for a category, its focus). Other stimuli are matched against it. As long as they are viewed as similar to it (S stands for similarity), the category expands its range. At level 1, R functions as the ground, S as the figure. Then, at level 2, the similarity is treated as known, i.e. the ground, against which difference (D) is juxtaposed: once stimuli start being viewed as different from the ones encountered so far, the span of the category is curtailed. The inherently mobile similarity is "fixated" at level 2 so as to function as the ground for introducing and emphasizing difference. The two figure-ground arrangements constitute a coherent whole, a vantage, which may consists of up to five such levels, usually three. However, a conceptualizer can only concentrate on one of the arrangements at a time, while the others recedes to his or her subconsciousness as presuppositions. The process of focusing on new coordinates against the background of old ones is called zooming in. The entailments in the right-hand-side column are aspects of actual behaviour, observable and measurable effects produced by the hidden cognitions.

Some colour data require a more complex treatment. For example, in some languages RED and YELLOW do not constitute separate categories but two types of vantage on the category WARM, called dominant and recessive. Figure 2 models the WARM category in Tzeltal, a Mayan language spoken in Chiapas, Mexico. The category is named with the dominant term k'an, focused in yellow, and the recessive term cah, focused in red:

Rozmiar: 62704 bajtów

Figure 2. WARM category in Tzeltal
(FC - fixed coordinates, MC - mobile coordinates; Y - yellow, R - red; S - similarity, D - difference)

In the dominant vantage, similarity is stronger (appears at more levels, which augments its value) and comes before difference. In the recessive vantage, difference is the stronger coordinate. As a result, that vantage's boundaries (margins) are established first and then whatever comes within those boundaries is seen as similar to the focal point. Due to the different arrangement of coordinates, the same mechanism of vantage construction yields different entailments.

The basis for coming up with this account were some 900 interviews conducted in Mesoamerica with speakers of 116 languages (plus many others elsewhere) with the help of the Munsell colour array:

Rozmiar: 62704 bajtów

Figure 3. Munsell array. © Hale Color Consultants, Inc. Used with permission.

The array is a set of 330 colour chips, 320 chromatic and 10 achromatic. The 320 chromatic chips are arranged in 40 columns according to hue and 8 rows according to brightness, while the 10 achromatic colours (from black through shades of grey to white) are placed to the left of the array in a single column. The array is a two-dimensional rendering of the Munsell solid and has to be severed for the purpose, which is usually done in the red area at column 40 or in the green area around columns 18-19.

The interviews involved three kinds of procedure: naming, focus selection and mapping. In naming, subjects are shown each of the 330 chips one by one in random order, and asked to name each. The results are then derandomized so that the area covered by each term can be marked. In focus selection, subjects are shown the full array without the naming ranges being indicated, and asked to mark the best example of each of the terms they have used in the naming task. In mapping they are again shown the "clean" Munsell array and asked to put a grain of rice on each chip they would name with a given term. When finished, subjects are asked to do so again, until they insist that no more chips can be named with that term. The same is repeated for the other terms.

In this way MacLaury identified three types of relationship between vantages: near synonymy, coextension and inclusion, plus the relationship of complementation obtaining between the dominant vantages of distinct categories. The relationships are found synchronically in world's languages but also follow a diachronic sequence in that order. The process has to do with progressively greater differentiation of categories along with greater emphasis placed on difference at the expense of similarity. The relationships are idealized segments of a continuum. The most intriguing of the four is coextension.

Other important VT constructs include:

    frame: if a category involves more than two, say three, vantages, they are grouped into frames; in Frame I, A is dominant relative to the recessive B, while in Frame II, B is dominant to the recessive C (C is thus ultra-recessive, but this term does not apply in a framed analysis);

    viewpoint: there are four types of viewpoint (VP-1, VP-2, VP-3 and VP-4), ranging from the most engaged or subjective to the most detached or objective; the two common types are the intermediate VP-2 and VP-3, while VP-1 and VP-4 are extreme cases of a continuum, probably unattainable to ordinary humans in normal categorization and language use

    stress: a notion related to viewpoint, connected with VP-2; it refers to the choice of coordinates (fixed or mobile) to which the conceptualizer attends more than to others and the degree to which this is done. In most languages, including Indo-European, fixed coordinates are stressed, but in some others one finds stress on mobile coordinates.


MacLaury, Robert E. 1995. Vantage theory. In Taylor, John R. and Robert E. MacLaury (ed.) Language and the Cognitive Construal of the World. [Trends in Linguistics. Studies and Monographs 82.] 231-276. Berlin/New York: Mouton de Gruyter.

MacLaury, Robert E. 1997. Color and Cognition in Mesoamerica: Constructing Categories as Vantages. Austin: University of Texas Press. [paperback 2011]

Why VT and linguistics?

In his review of Taylor and MacLaury's (eds. 1995) Language and the Cognitive Construal of the World, Michel Achard says:

[N]o specific theory has to date been able to account for both the linguistic and anthropological data. However, the convergence of insights and positions presented in this book allows us to hope that the two fields can very soon be considered together. In particular, it would be worth investigating whether Vantage Theory can be used successfully to describe linguistic data. (Achard 1999: 242)

Indeed, although Vantage Theory is not a theory of language, there is evidence that Achard's hopes are not futile. As developments in cognitive linguistics have shown, the science of language cannot stay isolated from findings in the fields of human cognition, psychology, perception and categorization. VT, although primarily a model of colour categorization, offers general insight into these spheres and at a few points specifically targets language. But first let us look at potential problems with the endeavour.

A few problems

Two kinds of transposition must be accounted for: (1) from categorization to conceptualization and (2) from colour domain to language. The former is partially discussed by Allan (2002) in his article on the English category of number:

[T]he grammar of number and quantification in English is exploited to reveal different points of view on the part of the speaker. The different points of view reflect different conceptualizations of what is spoken of... VT has hitherto been a theory of categorization, not conceptualization; but if VT is to apply more generally to language meaning, I believe it must tackle conceptualization. (Allan 2002: 684)

The linguist proposes his own version of VT, called VT2, and says:

If VT is a theory of points of view which give rise to categories, VT2 is a theory of points of view embodied in conceptualizations. (Allan 2002: 687; details of VT2 therein)

In the psychological literature, in turn, the distinction between categorization and conceptualization is often merely terminological. Some sources even define concepts in terms of categories:

A concept is a mental category that groups objects, relations, activities, abstractions, or qualities having common properties... The instances of a concept are seen as roughly similar; for example, golden retriever, cocker spaniel, weimaraner, and german shepherd are all instances of the concept dog. (Wade and Tavris 1990: 283; emphasis added)

Concept: A way of categorizing items and demonstrating which items are related to one another. In a concept-learning task, certain attributes of the stimuli are related to one another according to a specified rule. (EOP 1994, vol. I:284; emphasis added)

It seems that the parameters of hue, brightness and saturation can be seen as attributes, correlated with the rule of reciprocally balanced emphases on the similarities and differences between them.

In other sources it is possible to find formulations of striking similarity to the metalanguage of VT:

Concept generalization. The process whereby concepts are widened to subsume classes. In valid concept generalization, the individual notes points of likeness and difference among the separate objects or experiences, and then tests by the method of varying the concomitants to check upon the logic of the common thread running through them. (DOP 1947: 80; emphasis added)

The role of likeness and difference in the construction of concepts is thus parallel to that of similarity and difference in the construction of categories. I will assume on this basis that categorization is an aspect, a facet, or a mode of conceptualization. People conceptualize the world in many ways, one of which is a tendency to categorize it. (I owe this observation to a friend and colleague of mine, Przemek Łozowski of UMCS, Lublin, Poland.)

The second transposition, taking a step outside colour domain, requires specification of what might constitute the primary fixed coordinate for a conceptualization. In colour categorization the role is played by a hue (typically) or a degree of brightness or saturation (much less frequently). But what corresponds to these in linguistically encoded conceptualizations?

While several of the VT-and-language publications model language use in terms of vantages and by the same token provide an implicit answer to that question, none seems to do so directly. My tentative answer is as follows: the starting point for a conceptualization is a kind of rudimentary mental image arising in the mind of the speaker. Some support for this view comes from the author of VT himself, who while discussing categorization as space-time analogy, says that "a person constructs commensurable point of view on any experience in reference to images and to degrees of similarity and difference [...]". And further: "Images are equatable with [spatial, A.G.] landmarks because both are inherently fixed: they are discrete points that stand apart from other senses or ideas". If images can play this role in colour categorization (though, I admit, manifesting themselves as objectively measurable parameters of hue, brightness or saturation), they can certainly do so in (other) conceptualization. And indeed, in his comment on Taylor's (2003) paper on the coextensive relationship between high and tall, MacLaury (2003) says:

Vantage theory concerns the method by which people construct, maintain, and change categories. They do so by coordinating inherently fixed images (or ideas) with inherently mobile (or changeable) recognitions of similarity and difference of perceived experience to the images. [...] Vertical Extent (VE) is a lineal image of a relation between an upwardly positioned entity and the ground. By coordinating this image (VE) with attention to similarity (S) and attention to difference (D), the image becomes the reference point of a category. The viewer—the one who fashions this construction–can then make sense of (i.e. classify) any experience in terms of how similar and different it is to this image ... (MacLaury 2003: 285; emphasis added)

VT on language and cognition

General observations proposed by VT on the nature of language and cognition include the issues of subjectivity, the agency of the conceptualizer, and linguistic relativity/universalism.

Subjectivity has to do with the conceptualizer's point of view or perspective, notions without which cognitive linguistics would be hard to fathom. The two types of vantage are different points of view, different ways of conceptualizing the same entity or scene. VT proposes an explanation of these phenomena by referring to the variable degrees of attention to similarity and difference, the strength of their contribution to conceptualization, and the focusing, "stabilizing" role of inherently fixed coordinates. Subjectivity, then, is to be understood as a vital parameter of meaning, which "arises" out of the subject's dealings with the world, rather than being an inherent property of language units. Speakers (subjects) construe meanings within the bounds of their cognitive abilities.

These abilities both drive and limit the categorization-spacetime analogy. Within the limits, the conceptualizer enjoys a considerable amount of leeway (hence the multitude of category types in world languages) but operates according to the same fundamental cognitive processes. Thus, the conceptualizer is unconstrained in any dramatic sense by the language system he or she is using. In this sense VT is a non-Whorfian theory, but - paradoxically - one which stresses cultural and individual differences between speakers. For example, MacLaury found that two speakers of Tenejapa Tzeltal (Mayan family, spoken in Chiapas, Mexico) who have both spent their entire lives in the same village used the same colour terms but categorized the colours differently (cf. MacLaury 2000: 274-276). Relating these findings to the Sapir-Whorf hypothesis, MacLaury first quotes the famous excerpt from Whorf:

We are thus introduced to a new principle of relativity, which holds that all observers are not led by the same physical evidence to the same picture of the universe, unless their linguistic backgrounds are similar, or can in some way be calibrated. (Whorf 1956:214; quoted in and italics added by MacLaury 2000: 289)

and then proposes to replace the italicized part with

not even if their linguistic backgrounds derive from one rural hamlet. (MacLaury 2000: 289)

The agency of the conceptualizer operates within certain broad cognitive constraints to yield different results not because of the nature of the language spoken, but regardless of it. It is this cognitive flexibility that is responsible for intralinguistic variation:

Within constraints imposed by visual neurology, people shape their categories in accord with their inclinations to subject the world to broad or constricted points of view. This argument does not inveigh against relativity, which is shown herein to pertain cross-linguistically and between individuals. Rather, the claim is that language has no influence on the process. (MacLaury 2000: 276; emphasis added)

The view reconciles the universal nature of cognition with the diversity of categories and linguistic conceptualizations in world languages. Therefore, it should be attractive to cognitive linguists, who assume that meaning resides in conceptualization (Langacker 1991: ix) but also stress individual conceptualization patterns imposed by different languages.

More questions to answer

(Much light on many of these questions is shed by several publications listed in the VT Bibliography.)

Which areas of linguistics are especially open to analyses couched within the VT framework? Which ones are more problematic?

How to best understand a vantage? What analogues does it have in language? Can one provide clear and unambiguous linguistic examples of the dominant and recessive vantages? (The latter question due to Vyv Evans)

Should one strictly adhere to the "dominant-recessive" terminology? Should one preserve but expand or replace it? My proposals for names of vantage types used in various publications include: "a subdominant vantage" (also Allan 2002 in a different sense), "a default vantage", "a canonical/non-canonical vantage", "a micro-vantage", "a macro-vantage", "a focused micro-vantage", "a focused macro-vantage". None of these seem totally satisfactory for one reason or other. Preston (1993, 1994) augments VT with the processes he calls "panning", "suspension" and "pruning".

What relationship between vantages can be thought of (hierarchies, embedding, other)? (The question is due to Chris Sinha)

How does the notion of vantage as point of view relate to other accounts of the latter notion?

What other VT constructs figure as important in linguistic analyses?

Directions of further research

Modifications of VT may proceed in two directions. One would be an attempt to remain faithful to the details of VT and specify in what areas of linguistics it can be applied. This option runs the risk of getting lost in the labyrinth of the theory's complexities with little real progress or gain. The other direction would be to modify the details of the theory but endorse its general tenets, such as those pertaining to the agency of the conceptualizer or the issue of relativity. This, in turn, may lead to such departures from VT that warrant the use of other names of the proposed models (cf. Allan's VT2 or Głaz's EVT).

In my opinion the latter option is the more promising (although there is no shortage of theories "on the market"), as it gives priority to data. In this sense it follows the VT tradition, which arose precisely from its founder's dissatisfaction with existing models of colour categorization.


Achard, Michel. 1999. Review of John R. Taylor and Robert E. MacLaury (eds.) Language and the cognitive construal of the world. Berlin and New York: Mouton de Gruyter, 1995, 406 pp. International Journal of American Linguistics, 65-2. 240-242.

Allan, Keith. 2002. Vantage theory, VT2, and number. Language Sciences, 24, 5-6. 679-703.

DOP. 1947. Dictionary of Psychology. Philip L. Harriman. New York: Philosophical Library.

EOP. 1994. Encyclopedia of Psychology. Ed. by Raymond J. Corsini. 2nd edn. John Wiley & Sons.

Langacker, Ronald W. 1991. Concept, Image and Symbol. Berlin & New York: Mouton de Gruyter.

MacLaury, Robert E. 2000. Linguistic relativity and the plasticity of categorization. In Pütz, Martin and Marjolijn Verspoor. Explorations in Linguistic Relativity. 251-293. Amsterdam & Philadelphia: John Benjamins.

MacLaury, Robert E. 2003. Vantages on the category of vertical extent: John R. Taylor’s “high” and “tall”. Language Sciences 25, 3. 285-288.

Preston, Dennis R. 1993. The uses of folk linguistics. International Journal of Applied Linguistics, 3, 2. 159-259.

Preston, Dennis R. 1994. Content-oriented discourse analysis and folk linguistics. Language Sciences, 16, 2. 285-331.

Taylor, John R. 2003. Near-synomyms as co-extensive categories: “high” and “tall” revisited. Language Sciences 25, 3. 263-284.

Wade, Carole & Carol Tavris. 1990. Psychology. 2nd edn. New York: Harper and Row.

Whorf, Benjamin Lee. 1956. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf. Ed. by John B. Carroll. Cambridge, Mass.: MIT.

Created with Pajączek